Two-Dimensional Maximum Clustering-Based ScatterDifference Discriminant Analysis for Synthetic ApertureRadar Automatic Target Recognition

摘要

In this paper, a novel image feature extraction technique, called two-dimensional maximum clustering-based scatter difference (2DMCSD) dis-criminant analysis, is proposed. This method combines the ideas of two-dimensional clustering-based discriminant analysis (2DCDA) and maximum scatter difference (MSD), which can directly extract the optimal projection vec-tors from 2D image matrices rather than ID image vectors based on the cluster scatter difference criterion. 2DMCSD not only avoids the linearity and singular-ity problems frequently occurred in the classical Fisher linear discriminant analysis (FLDA) due to the high dimensionality and small sample size prob-lems, but also saves much time for feature extraction. Extensive experiments conducted on the moving and stationary target acquisition and recognition (MSTAR) public database demonstrate that the proposed method is more effec-tive than the existing subspace analysis methods, such as two-dimensional prin-cipal component analysis (2DPCA) and two-dimensional linear discriminant analysis (2DLDA).